The post below where I show that belief in the literal truth of the Biblical tends to correlate well with IQ scores from the General Social Survey on a denominational scale is getting a lot of response; enough of it is of low quality that I'll close the comment thread soon enough. As I observed the "truth" which I had extracted out of the data is rather banal; I doubt it surprised anyone that a "fundamentalist" attitude toward religious scripture tends to be associated with low cognitive ability. The correlation here is probably not one of simple causality in either direction. It seems the most plausible model is one which notes that various denominations tend to have particular socioeconomic profiles which shape a general cultural outlook. In the American South this was made most explicit, with a rank order of status from Episcopalians, to Presbyterians, to Methodists, to Baptists, and finally on down to marginal sectarians. These denominations tend to run in families, but, one may change denomination with relative ease in the United States in comparison to other nations. According to The American Religious Identification Survey reports that 16% of the adult population changes their affiliation during their lifetime. This level of churn is also probably not random; those who change their socioeconomic status may "trade up" or "trade down" in their church so as to feel more comfortable among their peers. I like to point out that the presidential candidate and wealthy lawyer John Edwards was raised a Baptist, but switched to Methodism as an adult. This is probably partially a reflection of his class status shift as well as his social liberalism. In contrast, though George W. Bush was raised an Episcopalian, he now worships in a Methodist church. This is a pretty good illustration of Bush's reinvention of himself as Middle American despite his patrician New England origins.
I doubt that the correlation between cognitive ability & unbelief in the literal truth of the Bible has that much to do with personal inspection of the Good Book. It is true that intellectuals within the Christian tradition, from St. Augustine onwards, offered that one had to be very careful in separating the allegory and metaphor from the literal truth. In contrast, naive literalism has tended to be a "low church" populist impulse. But I don't believe that most high IQ individuals pay inordinate close attention to the Bible, nor do I believe that low IQ denominations are necessarily steeped in scripture study (there is a wide variance among "conservative" groups which adhere to Biblical literalism in regards to the emphasis they put on scripture reading and analysis, from a general neglect among Pentecostals to a marinade among some Calvinist churches). What I think is going on is simply what we might term the Wisdom of the Crowds; people conform to the social and religious group which they identify with. Biblical literalism flourishes because most people trust pastors and parents who preach it. Similarly, a more metaphorical reading flourishes because authorities in other denominations reject fundamentalism. I do think that a deep reading of the scripture in their original languages as well as their historical context tends to erode a naive belief in the literal truth of the text. Mainline Protestant denominations and Roman Catholicism has relatively high educational standards for its clergy and theological professionals. At the other end of the spectrum many evangelical Protestant sects have no such requirement. The Assemblies of God is a good example of this phenomenon, in this sect higher educational experience can even be perceived as corrupting. There is a reason for this perception: education, wealth and acceptance does corrupt and assimilate. Methodism for example was originally an evangelical reform movement within the Church of England, but over the generations it has become thoroughly mainstream and tainted by modernism.
A chart which uses the demographic data from The Pew Religion Survey:
Note that if you remove Roman Catholics out of this the R2 goes up to 0.81. I think cultural context is a big prior condition in these sorts of generalizations; about 1/4 of Roman Catholics today are Latinos who have only recently left a religious-monopoly society. In those societies the Church has a pretty strong grip on the supply of religious ideas, and its top-down structure tends to dampen the availability of Biblical literalism as a choice among peers. In contrast, American Protestantism is very schismatic, every mainline denomination has a breakaway fundamentalist equivalent.
Note: I know that Jehovah's Witnesses and Unitarian-Universalists aren't Protestant, but they come out of Protestant culture.
Update: In response to John's comment below, I took the education data from The Pew Religion Survey and calculated the index of diversity. I've sorted it and normalized it below from the least diverse to the most (most would = 1). Most religious groups seem pretty diverse from what I can tell when it comes to educational level. Hindus are the least diverse, and Presbyterians the most, but it is interesting to note that conservative sects are also less diverse in terms of education. This would tend to agree with supply-side models which suggest that these sects are "high tension" and generally only those who are very proactive in identifying with these groups opt-in.
| Hindu | 0.85 |
| American Baptist | 0.85 |
| Church of Nazarene | 0.87 |
| Free Methodist | 0.87 |
| Independent Baptist | 0.88 |
| Assemblies of God | 0.89 |
| Southern Baptist | 0.91 |
| Evangelical Lutheran | 0.92 |
| Church of Christ | 0.92 |
| Lutheran Church, Missouri | 0.93 |
| Jewish | 0.93 |
| Episcopalian | 0.94 |
| Unitarian | 0.94 |
| Seventh-Day Adventist | 0.95 |
| Mormon | 0.95 |
| United Church of Christ | 0.95 |
| United Methodist | 0.95 |
| Catholic | 0.95 |
| Orthodox | 0.96 |
| Atheist | 0.97 |
| Disciplines of Christ | 0.97 |
| Presbyterian Church USA | 0.97 |
| Presbyterian Church in America | 0.98 |
- Log in to post comments
Another factor may be that Lutherans and Catholics are communitarian, multi-class churches that you are born into, and educated members don't trade out of the denomination (though they may switch congregations). Somewhat of a conjecture but I've read that these two churches tend to have the widest class range.
Just a comment that literalist churches agree abstractly on literalism, but disagree widely and violently on the details, and that many of them (e.g. dispensationalists) often do rely on metaphorical prophetic interpretations, for example of Revelation. No one really believes that a seven-headed dragon will rise from the sea. The literalism seems to be concentrated on specific docrines questioned by modernizers: 7-day creation, virgin birth, the life of Christ, and the miracles.
Higher education should be promoted
Anyone who has been a JW for decades like myself knows that college education was a defiant act which surely would get you sanctioned by the congregation.
We had a popular slogan called the "13 year plan" which was 12 years of public school which could include the vocational 'trade' schools,then one year of regular pioneering THEN you could get married.
The Watchtower religion is in dire straits for cheating millions of followers out of higher education with the "deliverance is at hand" slogan (still the 2006 district convention theme) the Watchtower is using apologist to discredit old timers who have been there.
Worldly secular statistics show the UU Unitarian universalism church as the most educated at 55% of members with a college degree and the Jehovah's Witnesses at 5%.
The Watchtower NOW has lessened up the restrictions as the 'new system' is on hold and the younger JW generation won't tolerate the janitorial jobs of their parents.
Although they have recently changed their policy so that at least a technical education is now seen as appropriate, the vast majority of Witnesses still have a high school education or less, and any change in their education level will take years before it has any effect on the Society
--
Danny Haszard born 1957 as a 3rd generation Jehovah's Witness
http://jehovahwitness.vox.com/
I was in the cult and now I'm out
The literalism seems to be concentrated on specific docrines questioned by modernizers: (...) virgin birth (...).
Actually this item is actually doctrine not Biblical literalism. It was agreed by bishops at some old council (Nicea?) and questioned by Luther precisely because it's not explicit in the Gospels.
That's an issue with Catholics (and surely Orthodox): they may be low in the Biblical literalism scale but high in the doctrinarian one. They believe more in the Church's doctrine than in the Bible itself. (Of course this only applies to true Catholics, not people who just vaguely identify with Catholicism but exerts independent judgement).
As I remember, there's a translation question about the meaning of a certain Hebrew word that was interpreted as a prophecy of Christ's birth. To those who accept the "virgin" translation, it's a literalist question. American literalists are probably much more literalist than Luther was; Luther was an associate of Erasmus, who produced a critical translation of the Bible from Hebrew and Greek which contradicted the Catholic Vulgate.
It's mostly liberal Protestants who question the virgin birth
on doctrinal questions "protestants" exhibit the most variance; for obvious reasons. the reformation ushered in the rebirth of unitarianism as well as more extreme 'primitive' forms of christianity.
This is an ecologic analysis, and as such is an extremely poor basis for drawing conclusions about association (to say nothing of causality!) at the individual level.
This is an ecologic analysis, and as such is an extremely poor basis for drawing conclusions about association (to say nothing of causality!) at the individual level.
if you know ABSOLUTELY NOTHING about decades of polling data which shows a negative correlation between years of higher education and fundamentalist religious beliefs, yes :=)
The IQ tests are weighted heavily on abstract thought (see Flynn), and post-graduate education also enhances or selects for abstract thinking.
Biblical literalism is a very concrete way of looking at the book, which would seem to be negatively correlated to either of these two factors (IQ and higher education).
It's too bad the IQ subtests aren't available to see if this "higher IQ" isn't actually a proxy for "more abstract thinking."
IQ: 150
Education Level: HS
Age: 21
Income: Too Damn Low
Spiritual Philosophy: Scientific Agnosticism
Theistic Beliefs: Deism (Impersonal Creator)
Religious Affiliation: Unitarian Universalism
Razib,
you seem to predict that if we isolated southerners, the correlation would go up. Is the data available to do that? (IQ, it seems, is not, but maybe education?) The diversity index you give might accomplish the same thing, but how to display it?
you seem to predict that if we isolated southerners, the correlation would go up.
i would bet on it; you'd remove more of the noise. i don't know of such data, but i'll look at some point. sure it's out there.
diversity index i just calculatd 1 - sum over all categories {(proportion of each category^2)}
there's a bunch of 'em
http://en.wikipedia.org/wiki/Diversity_index
Another source of similar data is:
CHRISTIAN SMITH, ROBERT FARIS (2005) Socioeconomic Inequality in the American Religious System: An Update and Assessment
Journal for the Scientific Study of Religion 44 (1) , 95-104
From a different angle though, the type of education and area of specialty varies along with religiosity. So highly educated people in the social sciences are the least religious and highly educated people in the purely technical and applied fields are relatively more religious. So it is more likely that a highly educated member of a fundamentalist denomination has recevied that education in a relatively narrow field such as engineering rather than, say psychology.
Neil Gross and Solon J. Simmons. 2006 "How Religious are America's College and University Professors" in Douglas Jacobsen (ed) The Post Secular University Oxford, Oxford University Press
According to the recent RAAS (Religious Attitudes Among Scientists) study, social scientists and natural scientists in the U.S. are very similar in terms of (ir)religiosity overall, with more variation from field to field within one or the other than between the two.
Over half are irreligious, and something like a quarter buy into orthodox religion. (Three fourths will say that the Bible is a book of fables, for example.)
Non-scientists in academia may be more religious. IIRC from other sources, engineers and physicians are more religious than scientists, but philosophers even less religious.
That paper is downloaded about for free (pdf) from Blackwell:
http://www.blackwell-synergy.com/toc/jssr/44/1
Most JSSR issues cost money, but that one's free.
As I remember, there's a translation question about the meaning of a certain Hebrew word that was interpreted as a prophecy of Christ's birth. To those who accept the "virgin" translation, it's a literalist question. American literalists are probably much more literalist than Luther was; Luther was an associate of Erasmus, who produced a critical translation of the Bible from Hebrew and Greek which contradicted the Catholic Vulgate.
It seems you are right. I must have misplaced that with some other creed item. My bad.
I have a post up at my blog about some of the issues with graphing percentages and fitting lines, with a plot of my own that I think makes the relationship in the Postgrad Education vs Biblical Literalism case much clearer.
http://ecstathy.blogspot.com/2008/05/relationships-between-percentages-are.html
Note that if you don't make the (to be frank) crazy assumption that the relationship is linear, the "Roman Catholic" point is not in any sense an outlier. In my version of the graph, it's clearly in keeping with the general relationship.
You included data for the PCA (Presbyterian Church in America) in your diversity index. Do you not have IQ data for it? How about the Missouri Synod of the Lutheran church?
I checked the actual Pew survey. They actually did distinguish between conservative denominations and not, though in the Presbyterian tradition, the left out the OPC, which is older than the PCA. They also left out the CRC's conservative counterpart - the URCNA. Either way, there did appear to be fewer postgraduate-educated people by percentage in the Missouri Synod and PCA than the mainline Lutheran tradition and the PCUSA, respectively.
You included data for the PCA (Presbyterian Church in America) in your diversity index. Do you not have IQ data for it? How about the Missouri Synod of the Lutheran church?
right, no IQ data.
Yes there seems to be higher religiosity in technical as opposed to theoretical disciplines. For example, equally educated professors who are engineers are more religious than physicists, social workers are more religious than sociologists, therapists are more religious than psychologists. The least religious disciplines are those dealing with knowledge that is self reflective, that is concerining one's own thought processes: philosophy, psych, soc, anthro. I have observed that among psychology professors, the least religious are the social psychologists and the most religious are clinical and organizational psychologists (again, the most applied disciplines).
So if i were to make a generalization, it appears education that teaches critical thinking about ones' own thought processes appears to be associated with the least religious individuals. But education that is primarily concerned with a set of concrete knowledge or skills external to the individual appears to be less corrosive to traditional religion.
So if i were to make a generalization, it appears education that teaches critical thinking about ones' own thought processes appears to be associated with the least religious individuals. But education that is primarily concerned with a set of concrete knowledge or skills external to the individual appears to be less corrosive to traditional religion.
if you make social science generaliations it is appropriate to cite a study (i can name some, but i'm not going to do your work for you) if you want to be engaged with. otherwise, you're just farting.
By the way, I got the name wrong before---it's "Religion among Academic Scientists," for your Googling pleasure.
Be very careful, though, about how the results are being discussed. Ecklund and Scheitle seem to want to spin things to suggest that advanced scientific knowledge doesn't erode religious belief. (For example, saying that becoming a scientists doesn't "necessarily" make folks just "drop" their religion, without being clear about how much it tends to make them less religious or less orthodox.)
This isn't too surprising given that the research is funded by the Templeton Foundation (and NSF, I think).
Many people reporting on it want to make it sound like "most scientists aren't atheists" rather than "most scientists aren't theists." Really, most scientists are atheists in the broad sense, i.e., nontheists.
E&S make it sound like the main thing that accounts for most scientists being nontheists is self-selection out of science by theists, not decreasing theism among people who do go into science. Even if the former effect is bigger, I suspect the latter is not small.
By their measures atheists and agnostics are only 8 percent of the general population, but about 60 percent of scientists at the fairly elite public and private universities they studied.
In that tier of academic science, it seems, you have 7.5 times as many atheists as in the general population.
E & S want to make it sound like the "best predictor" of a scientists religious beliefs is religiosity in the household they were raised in. I think they're looking through the wrong end of the telescope. Even if they were right that science doesn't seriously erode religious belief---which I doubt---it's striking that by far the "best predictor" of scientific achievement is atheism.
The usual suspects of race, sex, graduate education, and high family income don't come anywhere close to being as good a predictor of scientific achievement. (The large majority of well-off folks in the US are theists; so are most white males with graduate degrees.)
If you look at Larson and Witham's study of NAS members, the results are even more striking, with 93 percent nontheist, and only 7 percent theists. Apparently 8-14 percent or so of the population is doing 93 percent of the top science. (There's some iffiness in survey measures of nontheism, but it doesn't matter much.)
Non-theism is a vastly better predictor of top-notch scientific achievement than the usual suspects. Nontheists are about 100 times more likely to be top scientists than theists, and outright atheists are even more likely. (Most nontheists in the general population are non-believing "agnostics," not disbelievers; most NAS members are outright disbelieving "atheists.")
If there's a "godless elite," it's the scientific elite, not business or the media elites, or even the educational elite.
I think this is the most under-reported angle in the whole science-and-religion/elites thing.
Many thanks for the stimulation and a sort of piggy-back blog based on your work from another with an interest in denominations and religion. See
http://pluralistspeaks.blogspot.com/2008/05/effect-of-postgraduate-educ…
My distant wife is doing an MA statistics course and I asked her to comment but she has a strong relationship between time of day and sleep, very close to 1. My own understanding is visual, and like your commentator I used to draw lots of curves showing marginal relationships - utility and indifference.
Here's an online working paper version... I don't know if it's much different from the book chapter you cite:
http://www.wjh.harvard.edu/soc/faculty/gross/religions.pdf
It gives a breakdown by eliteness (from community colleges to elite Ph.D.-granting institutions), but not by discipline.
Here's Ecklund's working paper "Initial Findings from the Study of Religion Among Academic Scientists (RAAS)"
http://www.epi.elps.vt.edu/Perspectives/ecklund.pdf
It doesn't do a breakdown by discipline, either. :-(
Um ok generalizations based on studies. we'll see who is farting:
1) the Leuba survey of NAS members which was replicated by Larson and Witham mentioned above found the following breakdown among types of greater scientists (NAS members) to "belief in god": physical scientists: 35%, biological scientists: 17%, sociologists: 19%, psychologists: 13%. this general trend was reflected in his non elite sample as well.
2) in Otis and Alcock's surveys of critical thinking:
Alcock, J. and L. P. Otis. (1980) Critical Thinking and Belief in the Paranormal Psychological Reports, 46, 479-482.
Otis, L. P. and J. E. Alcock. (1982) Factors Affecting Extraordinary Belief. The Journal of Social Psychology, 118, 77-85.
psychologists were always the least religious and spiritual academic discipline. Mathematicians, physicists, zoologists, etc were always higher.
3) in the Gross and Simmons study mentioned above, the most atheistic disciplines were psych and bio (at around 60%) with mechanical engineering more around the 50% mark and comp sci around 40%.
4) in a survey by Public Perspective in 1991, the % of academics marking "religion = none" was highest in antro at over 60%, then philosophy, psych, and soc at over 50%. Bio, math, physics and chemistry were all lower than 40% religion = none.
I'd say thats sufficient evidence to make that generalization, unless you have something contradicting it.
It's a poor substitute for the original paper, but here's a magazine article and a newspaper article about the Gross & Simmons research, including a few stats about disciplines:
http://harvardmagazine.com/2007/07/faculty-faith.html
http://www.nysun.com/new-york/professors-find-god-in-groves-of-academe/…
Watch out for some of the generalizations about social science. The RAAS did not find a big difference between natural scientists and social scientists at elite institutions.
IIRC, RAAS found that among natural scientists, chemists are more likely to believe in god than biologists or physicists, and political scientists are more likely to believe in god than some other social scientists. The difference between fields within natural or social sciences was bigger than the differences between natural and social sciences overall.
Trying to put Gross & Simmons together with Ecklund, I'd say that
1. G & S show that the more elite the institution, the less theistic the professors. (The NAS data from Larson and Witham seem to indicate that the elite science / atheism correlation goes all the way up.)
2. G & S show that in you find more believers in applied fields like Accounting, elementary education, finance, art, criminal justice, and nursing than in core science fields.
3. The RAAS didn't find a big difference between natural and social sciences, but did find differences from science to science.
4. The RAAS press releases seem to want to make it sound like scientific education doesn't erode religious belief, but they waffle around it in a suspicious way. Even if it doesn't, that leaves a huge unexplained correlation between irreligion and scientific achievement. Maybe atheism causes science, or religion kills it, or something else erodes religion and promotes science. Whatever it is, it isn't mostly the usual suspects of correlations with age, sex, and income.
My impression is that engineers are more likely to be believers than scientists. It surprises me that Gross and Simmons found mechanical engineers especially likely to be unbelievers. That may be mostly because they were comparing them to a broad variety of fields, not mostly scientists.
I seem to recall seeing some stats showing that indeed engineers are more likely to be religious than (natural?) scientists, but I forget where.
I would not buy Gross's argument that family (of origin) income has a lot to do with unbelief among professors. Family income is positively correlated with being atheistic, and with being a scientist, but the effect is nowhere near strong enough to explain why so many top scientists are atheists; there's a bigger factor left unexplained.
It's not just biblical literalism that's inversely correlated with advanced scientific achievement. Atheists outperform agnostics, who outperform liberal theists, who outperform orthodox theists. Religious orthodoxy seems to be strongly inversely correlated with scientific achievement across the scale.
(This thread at Framing Science has more on that, including some links to stats about income and atheism, race and and education effects, etc.:
http://scienceblogs.com/framing-science/2008/04/does_advanced_science_e…)
Here's a link to a .pdf of a paper by Ecklund about the RAAS. Not the best paper for our purposes, but it's free:
religion.ssrc.org/reforum/Ecklund.pdf
Among other things, she notes that among the "elite" university scientists she studied, "fundamentalist" or "evangelical" christians are especially rare.
religion.ssrc.org/reforum/Ecklund.pdf
On general principles here's the short Larson & Witham paper about their NAS results, from Nature:
http://www.stephenjaygould.org/ctrl/news/file002.html
Note that not only are 93 percent nonbelievers, 72 percent are outright disbelievers; given that in the general population more people are agnostics than atheists, atheists are clearly outperforming even agnostics at that level, by a factor of something like 5.
I suspect that not only do nontheists outperform theists by a factor of around 100 at the NAS level, but that atheists outperform evangelicals by a factor of around 1000 or so. (But the n for evangelicals in the NAS may be so small that the statistics would be iffy.)
I got ahold of the actual Ecklund & Scheitle paper, "Religion Among Academic Scientists: Distinctions, Disciplines, and Demographics."
A quick look at their Table 4 suggests that they're overplaying the selection-going-in and downplaying the erosion of religious belief by science. Yes, the selection effect is huge---strongly and orthodoxly religious people generally don't go into science, or don't make it anywhere---but even accounting for that, science and religion are inversely correlated.
Table 4 shows that most U.S. scientists were raised Catholic or Protestant, but almost three fifths stop identifying as Catholic or Protestant. Likewise the percentage of "no religious affiliation" almost quadruples.
The incidence of atheism and agnosticism confirms the trend; despite most being raised in at least somewhat religious households, about 73 percent are atheist or agnostic, with another 7 or 8 percent believing in some "higher power" that they say is not God.
The way E & S frame it, religiosity of the household of origin may indeed be the "best predictor" of religiosity among scientists. But still, science itself is a great predictor of irreligiosity, even among scientists raised religiously. The large majority of scientists is raised in religious households---albeit disproportionally theologically liberal ones---but we end up overwhelmingly irreligious.
E & S are right that becoming a scientist doesn't necessarily cause you to "simply drop" your religious commitments---but who really thought that it did?
Their own statistics seem to confirm what they're trying not to admit; while selection going into science may explain more of the huge correlation between science and irreligion, the erosion of religious belief seems to be very strong as well.
paul, can i get a copy of that? contactgnxp@gmail.com
tx
oops... bobbled a statistic in my last comment. The 73 percent atheists and agnostics is if you INCLUDE the 7 or 8 percent who believe in a higher power that's not God.
In more detail, here's the percentages of natural scientists (NS) and social scientists (SS) who chose the various answers on the God quiz:
"I don't believe in God" 37.6 NS, 31.2 SS
"I do not know if there's a God and there's no way to find out" 29.4 NS, 31.0 SS
"I believe in a higher power, but it is not God" 8.2 NS, 7.2 SS
"I believe in God sometimes" 4.2 NS, 5.4 SS
"I have some doubts, but I believe in God" 12.9 NS, 15.5 SS
"I have no doubts about God's existence" 7.8 NS, 9.7 SS
If we only count the last two as "theist" answers, we get 20.7 percent NS and 25.2 percent SS as theists. If we count the last three (including believing only "sometimes") we get 24.9 percent NS and 29.6 SS.
So about a quarter or so of these scientists are theists, but less than 30 percent. If you add in the ones who believe in a higher power that's not God, you get about a third being even vaguely religious in belief terms.
Practice-wise, over half didn't go to church at all (55.3 SS, 50.3 SS), and about another quarter didn't go more than 5 times a year (23.1 NS, 26.5 SS), so over three fourths go at most a few times a year. (Maybe for cultural/social reasons, not seriously going to church.) At the other end, about 14 percent go to church twice a month or more (12.9 NS, 14.4 SS).
Razib,
Can you plot actual distributions of education level or IQ, as opposed to just the means?
If you lumped the denominations together into three or four clusters by degree of orthodoxy/fundamentalism, and then superimposed the actual plots, that could be very enlightening.
In particular, it might show that the upper tails are just too low for more fundamentalist folks to make it in science. To be a scientist, you pretty much have to have an IQ in the upper quartile or maybe decile. (Whether IQ is a good measure of "general" smarts or not, it's fairly predictive of academic success.)
Means don't matter as much as the actual numbers/percentages of people who are well above average, so seeing the tails would be more interesting than seeing the means.
A plot like that might show that IQ alone is a sufficient explanation of the huge correlation between irreligion and scientific accomplishment, without taking religion into account. Or it might not. Either way, it'd be interesting.
BTW, I wouldn't be surprised if the IQ and education distributions by denomination were relatively narrow.
To oversimplify, it may be that almost everybody with an IQ more than a standard deviation or two above the mean for their denomination tends to find a "smarter" denomination where they're not such an outlier.
If so, any denomination with a mean IQ a standard deviation below the population mean would have almost no members with IQ's more than a standard deviation above the population mean. (And any denomination whose mean IQ was not substantially above the population mean wouldn't have many members more than 2 SD above the population mean, as is probably usual in elite science.)
Those made-up numbers are kind of bogus, because the range of denomination means isn't really 2 SD, but you probably get the idea.